You meant march? Anyway I lost track because the 31 March patch could be called either the March or April update 11
Goths have high winning rate but low pick rate, so most of their wins are merely random civs encounters.
This proves my point in the previous post, even if this are 1x1 stats, both persians and Khmer have such a high pick rate, i understand that for persian on mixed maps, but khmer lol
Yes, I mean March, i got my months mixed up.
At 1600+ ELO, Khmer only have a 50.2% win rate, while Persians have a 52.4, so neither is very OP at all, at least for the best rated players.
Celts do much better than both at 54.4%, and most people would never say Celts are OP.
Sure itâs not 33/33/33 between Chinese/Persian/Aztec, but Persian play rate is quite inflated, likely because itâs the most âflashyâ civ after DE buffs, and there is so much discussion about their OPness.
Well, since often people who play random end up against a civ picker, it means these Goth players had to fight quite a lot of Chinese/Persian/whatever. So itâs still impressive.
I see that only for Persian. Khmer are less picked than Franks and Huns for instance if we choose âall ELOâ and they are as popular as Japanese with â1600+ ELOâ, while being less popular than Aztecs and Mongols.
Imo the 1600+ is just not enough separation, it could be 1600-1800 and 1800+ or 1600-2k and 2k+ or 1600-1800, 1800-2k and 2k+.
I think it serves itâs purpose, hoever. Could be more detailed, but it is a slap of reality on all our faces.
Well yes surely this is better than no info and guesswork mostly until now.
Man, I still cannot believe Goths, Celts and Indians are top tier OP GG civs!
Franks were always dominant, but Celts and Goths? I guess Infantry and Camels are stronger than we knew.
Perhaps Elephant archers are OP 1v1 units with few counters too.
Goths winrate could also be a result of people getting it at random, against another random who was even a worse civ. Pickrate is still just 1% so itâs easier to have a higher winrate with sheer luck of the statistics, where as a civ that is truly picked alot (ie. pers) must be a good civ if it has also a good winrate. But indeed I would like a more separated higher tiers to see even clearer whatâs happening at each level, but for now this is very good information nonetheless. (again, if this can be trusted)
That they get picked so little, yet win so much, is a sign that they are much stronger than assumed, however.
Persians get so much âthey are broken OPâ publicity, yet they are actually very balanced, and have a worse win rate than Celts.
Celts is a civ with only 2 Imp Age Blacksmith upgrades, very poor Archery Range and Stables. I would never assume they are OP, but the evidence seems to point that way.
Man aztks and japanese have their own maps, arabia and arena are maps for the first, while mixed maps are a japanese pick, it is not like they pick those civs on any map, mongols are a 90% megarandom civ. You need a better understanding of how and why players pick those civs, the winning rate is another story.
The fact that khmer have such high pick rate it only shows how strong the civ is despite not having a map that boost their attributes like persians, aztk, mongols, etc.
Now talking about winning rate, franks have the highest winning rate since age of rajas and despite that no one actually picks them in tournaments unless the map has tons of bushes which already happened and was changed on nilis tournament, now they pick mongols on that map lol.
Arenât they supposed to be OP on closed maps where they have an easier time to use their farm bonus and skip building bonus?
The deal is that everyone was well aware that Franks were the best before DE, and we had aoestats to verify it. But after DE not only such data wasnât available anymore, but the pathing proved quite hard to handle, so it was logical for people who wants to win to try their luck wth archers instead.
Btw when I brought up the 1v1 tier list in the Khmer thread you replied you were complaining about team games, and now that you think it supports what you think 1v1 data is relevant again? Thatâs not how it works.
Since only 1% of players are above 1600, making 4 tiers with 0.25% of the player base each would make little sense (and I guess actually the repartition would be worse)
Well, since they were supposed to be either the worst or only above Turks I donât see that happening often⊠Also, other objectively better civs were just as rare to encounter (like Viet or Lithuanian) and yet they failed to accomplish such a high win rate, so I donât think sheer luck can explain this placement alone (but itâs still sensible to remember about a potential bias, sure)
This is pretty much what I was trying to say. It could be that it is showing exactly as things are indeed.
So Goths need nerf and Franks 20 knights in 20 minutes rush OP
11
Portugese, Saracens, Koreans all bottem tier for all ELO ranges.
Now lets react seriously:
My bet they used aoe2.net api to gather data and some how managed to determine who win/lose the game. This means they can expand this site with win rates for every map and for teamgames. That would be great!
They can even add DM or unranked at some point i time. I very much like this page! If what i say is true, it is also possible to make winrates by civ by player, like voobly has. That would be cool too!
I see many great opportunities! Good job for the creator!
For winrates by civ and map for individual players check out aoe2.club.
I guess that it wont take long until we see more detailed statistics for all winrates (per map, more elo ranges)
they used to have arbalest when they started and thus a full archery range tree
So based on the website for all ELO categories, and assuming that weâre aiming for a 55%-45% win rate for all civs, then Saracens, Koreans, and Portuguese need buffs?